[KAFKA-540] log.append() should halt on IOException

• Type: Bug
• Status: Closed
• Priority: Blocker
• Resolution: Fixed
• Affects Version/s: 0.8.0
• Fix Version/s: 0.8.0
• Component/s: core
• Labels: bugs
• Assignee: Jun Rao
• Reporter: Jun Rao
• Created: 03/Oct/12 22:34
• Updated: 04/Oct/12 03:47
• Resolved: 04/Oct/12 03:47

Description

See the following entry in the broker log in a system test run. We interrupted the ReplicaFetcherThread during shutdown. However, log.append halts the system when we hit the interrupted exception. The fix is not to halt the system in log.append and just pass on the exception. The caller can decide what to do.

[2012-10-03 15:08:53,124] FATAL [Kafka Log on Broker 2], Halting due to unrecoverable I/O error while handling producer request (kafka.log.Log)
java.nio.channels.ClosedByInterruptException
at java.nio.channels.spi.AbstractInterruptibleChannel.end(AbstractInterruptibleChannel.java:184)
at sun.nio.ch.FileChannelImpl.write(FileChannelImpl.java:203)
at kafka.message.ByteBufferMessageSet.writeTo(ByteBufferMessageSet.scala:128)
at kafka.log.FileMessageSet.append(FileMessageSet.scala:155)
at kafka.log.LogSegment.append(LogSegment.scala:60)
at kafka.log.Log.liftedTree1$1(Log.scala:282)
at kafka.log.Log.append(Log.scala:270)
at kafka.server.ReplicaFetcherThread.processPartitionData(ReplicaFetcherThread.scala:42)
at kafka.server.AbstractFetcherThread$$anonfun$doWork$5.apply(AbstractFetcherThread.scala:105)
at kafka.server.AbstractFetcherThread$$anonfun$doWork$5.apply(AbstractFetcherThread.scala:98)
at scala.collection.immutable.Map$Map1.foreach(Map.scala:105)
at kafka.server.AbstractFetcherThread.doWork(AbstractFetcherThread.scala:98)
at kafka.utils.ShutdownableThread.run(ShutdownableThread.scala:50)
(